Lancet study now available on web

posted at 10:15 am on October 12, 2006 by Allahpundit

Tony Blair and John Howard say it’s nonsense. Jan Egeland, the UN’s Undersecretary-General for Humanitarian Affairs, is “surprised” by the findings. But the nightly news shows can’t get enough.

Have at it, number-crunchers. Note that Baghdad appears to be relatively safe compared to other parts of the country.

lancet.png

More (Bryan): The Lancet study would have us believe that 2.5% of Iraq has been killed by the war in the past three years. It would have us believe that more Iraqis have died as a result of a mid-sized insurgency than Americans died in World War II. Or the Civil War. Or Germans, who died in World War II, fighting against the combined might of the USSR, the British Empire and the United States, at a time when Germany was reduced to conscripting young boys and old men to resist those armies as they approached Berlin.

This study, in other words, is nonsense on stilts.

And further, where are the bodies?

In science, extraordinary claims require extraordinary proof or evidence. The Lancet study offers neither. It was extrapolated from a handful of deaths in ways that, imho, don’t apply to unnatural causes like war. It’s nonsense.

Update: Omar at Iraq the Model responds to Lancet.

Update: Megan McArdle is skeptical:

The average report is of about 30 civilian casualties a day–a horrifying number that should sting the consciences of those who advocated war. I’m sure that there are more whose deaths go unreported. But assuming that violence is the major cause of death, how likely is it that the newspapers are all consistently underestimating the number of violent civilian deaths by a factor of five or more? Okay, maybe they’re all happening outside of Baghdad. Except outside of Baghdad includes the Kurdish north, where 10% of the population is mostly not getting shot by insurgents. And a lot of Iraq’s other towns outside of Baghdad aren’t that big. Bayji, a major oil centre, has 60,000 inhabitants. As anyone from a town that size can tell you, it wouldn’t go unnoticed if it was losing 1,600 people a year to murder.

Update: We’ve gotten a few trackbacks today from left-wing assholes asserting that we’re mindlessly dismissing the survey, even while they go about mindlessly asserting it’s accurate. Of course, we haven’t (or at least, I haven’t). In fact, my point yesterday in our post about this was that the methodology probably isn’t all that suspect, and whatever the actual numbers are, they’re far too many.

The Commissar is thinking along the same lines and has a post worth reading in full. Like me, he thinks the numbers are exaggerated, but not entirely. And he’s got a theory as to how — and why.

Update (Bryan): Forget what I said about German casualties in WWII. I had the numbers way wrong. But the rest still stands, and I think the Lancet study is extremely dubious.

Breaking on Hot Air

Blowback

Trackbacks/Pings

Trackback URL

Comments

Applying Lancet’s methodology, the Democrats will get 95% (including excess votes) come November.

Much like Saddam did.

Niko on October 12, 2006 at 10:23 AM

Yes, the old medieval style of counting. It feels like we fought 259,000 English types. History finds it is actually 10,009. So we’ve come to this.

Hening on October 12, 2006 at 10:29 AM

I’m not a statistician or an economist (thank you, God) so I can’t speak to the numbers, but the method of obtaining the sample does not seem truly random. Anytime you allow human decision on the spot to interfere with a random sampling method, selection bias can come into play. That does seem to be the case with this study:

Decisions on sampling sites were made by the field manager. The interview team were given the responsibility and authority to change to an alternate location if they perceived the level of insecurity or risk to be unacceptable. In every cluster, the numbers of households where no-one was at home or where participation was refused were recorded. In every cluster, queries were made about any household that had been present during the survey period that had ceased to exist because all members had died or left. Empty houses or those that refused to participate were passed over until 40 households had been interviewed in all locations.

Wandering the streets until you find a particular number of represetative opinions is not the ideal method of finding a random sample, so it seems to me there could be some selection bias present on the part of those who conducted the study. Would it be enough to skew the results? Well, if you extrapolate from a bad sample, then you get bad data. I think they got a bad sample, becuase their methods for ensuring a random survey were not great – due either to researcher sloppiness or physical constraints. I think it’s the latter.

The researchers should have also looked at quantitative data separate from what they got from the interviews (death certificates). They could have done a data mine of information from hospitals, morgues, funeral homes and mosques to get numbers on funerals and autopsies. This data could have been used in the final analysis and if it had, probably would have limited criticism of what they did produce, since the data sample would likely be stronger.

Slublog on October 12, 2006 at 10:39 AM

Hard to believe this many people have died. On the other hand, hard to believe both John Hopkins and Lancet would participate in shoddy research, this is their area of expertise after all. Curious.

honora on October 12, 2006 at 10:49 AM

Gobbledegook + wishful thinking + BDS + flawed technique = Dinosaur media orgasm

SouthernGent on October 12, 2006 at 10:52 AM

I wouldn’t call it gobbledygook. The numbers are almost certainly wrong, but the actual number is probably a lot higher than Iraq Body Count has it.

Allahpundit on October 12, 2006 at 10:55 AM

Doesn’t the timing all by itself disqualify the study? Just as the one before the presidential elections?
I mean, The Lancet has a track record when it comes to this kind of stuff.

Kim Hartveld on October 12, 2006 at 10:58 AM

I wouldn’t call it gobbledygook. The numbers are almost certainly wrong, but the actual number is probably a lot higher than Iraq Body Count has it.

IBC has the number at around 50,000. What do you think it is? My best guess would be somewhere between 125,000 and 175,000, if you take into account the secondary effects of the invasion.

Slublog on October 12, 2006 at 10:58 AM

Maybe even more. Whatever the number is, it’s a lot more than I expected.

Allahpundit on October 12, 2006 at 11:01 AM

Bryan, What do you mean here?

Or Germans, who died in World War II, fighting against the combined might of the USSR, the British Empire and the United States, at a time when Germany was reduced to conscripting young boys and old men to resist those armies as they approached Berlin.

Germany lost 7 million people in WWII. I think I read some comparison to Iraq and the number of Germans “killed by Allied aerial bombardment.” But even that might be more than 600,000.

But as it stands, that item needs work.

commissar on October 12, 2006 at 11:02 AM

From the press release from Johns Hopkins. Couple of interesting items in the release:

Males aged 15-44 years accounted for 59 percent of post-invasion violent deaths

Does that profile look familiar?

The proportion of deaths attributed to coalition forces diminished in 2006 to 26 percent. Between March 2003 and July 2006, households attributed 31 percent of deaths to the coalition

So where are the other 70-75% of the deaths coming from? The researchers certainly wouldn’t want the world to know that they could be related to the terrorists that don’t exist in Iraq, would they? Or how about victims of Saddam’s regime that were dug up from mass graves and identified ater the invasion? And this statistic alone would bring their numbers of “war dead” down to around 170,000, even if the numbers were accurate.

TexasRainmaker on October 12, 2006 at 11:03 AM

IBC puts it at about 50K, Brookings (IBC+UN) at 60K, and the LA Times puts it at 50K, but estimates that is undercounted.

I’m guessing 80K-100K.

Bob Owens on October 12, 2006 at 11:06 AM

While I think the Lancet numbers are too high, I think the percentage of “unexplained” deaths can be partially attributed to the fact that a country’s mortality rate tends to increase at times of violent conflict for a number of reasons.

People die in increased numbers because supplies may be unavailable, living conditions may be unsanitary, accidents happen with unexploded munition or the rubble of war. These deaths may not be directly a result of conflict, but they are a result of the war.

Slublog on October 12, 2006 at 11:07 AM

So where are the other 70-75% of the deaths coming from?

Sectarian violence: Shiite on Sunni, Sunni on Shiite. I thought this was pretty well documented at this stage? Whether these are properly called terrorists is beyond me. I believe the issue you are referring to is the belief that there are relatively few foreign terrorists in Iraq.

honora on October 12, 2006 at 11:11 AM

I’m also not a statistician, but if you’re going into neighborhoods and asking groups of families the same sorts of questions, wouldn’t they all give the same sorts of answers? If all 40 families in a neighborhood said they knew the same 50 people who were killed, is that 50 fatalities for the study, or 2000?

High Desert Wanderer on October 12, 2006 at 11:16 AM

Hard to believe this many people have died. On the other hand, hard to believe both John Hopkins and Lancet would participate in shoddy research, this is their area of expertise after all. Curious.

I think part of the bad research can be attributed to the fact that a cluster survey method was used. This is a common medical research method, ususally used to determine the incidence of a particular disorder in a scattered population. I believe it was improperly used here, since it works best with homogenous groups and like clusters. I don’t believe the rather diverse population of Iraq meets that criteria, but I think they used it because it has the advantages of being inexpensive and easy to use on a large area of land.

Slublog on October 12, 2006 at 11:16 AM

I’m also not a statistician, but if you’re going into neighborhoods and asking groups of families the same sorts of questions, wouldn’t they all give the same sorts of answers? If all 40 families in a neighborhood said they knew the same 50 people who were killed, is that 50 fatalities for the study, or 2000?

High Desert Wanderer on October 12, 2006 at 11:16 AM

Caveat: I have not read the study but from what I have read, the methodoly involved asking HH for deaths in their HH and getting death certificates for about 75% of reported deaths. Presumably these would have been checked for duplicate.

honora on October 12, 2006 at 11:24 AM

I ask … why do we care?

Discounting the victims of the terrorists and insurgents themselves … I say the higher the number, the better. The high number seems to be portrayed as a BAD thing.

Remember that this is not a war against 18 innocent kids as most previous wars have been. For the most part, these are radical insurgents and terrorists who if not killed now … would surely kill others at a later date – war or no war.

The “War on Terror” can never be complete if even one “terrorist” is left alive.

Gregor on October 12, 2006 at 11:26 AM

Hard to believe this many people have died. On the other hand, hard to believe both John Hopkins and Lancet would participate in shoddy research, this is their area of expertise after all. Curious.

I think part of the bad research can be attributed to the fact that a cluster survey method was used. This is a common medical research method, ususally used to determine the incidence of a particular disorder in a scattered population. I believe it was improperly used here, since it works best with homogenous groups and like clusters. I don’t believe the rather diverse population of Iraq meets that criteria, but I think they used it because it has the advantages of being inexpensive and easy to use on a large area of land.

I have experience with market research and some reseach methods. I also understand the drawbacks of the methodology used–assume this is the reason for the wide variation in the result. Nonetheless (no disrespect intended) I would need more evidence that the method yielded bad data, being somewhat familiar with the reputation of both institutions.

honora on October 12, 2006 at 11:30 AM

“On the other hand, hard to believe both John Hopkins and Lancet would participate in shoddy research, this is their area of expertise after all. Curious.”

not picking on you, honora, I too used to take studies that came out of various “prestigious” institutions as gospel. after all scientists would never let an agenda corrupt the analysis of raw data. Yes fundamental toolie flaws in the analysis used might exist but NEVER EVER would an agenda corrupt the process.

In today’s world unfortunately I know not whom to trust and “prestigious” returns to it’s archaic definition, (thanks m-w.com)…

1 archaic : of, relating to, or marked by illusion, conjuring, or trickery

I revert to, if it sounds to good to be true than it is not. An axiom, paraphrased courtesy of South Park, if it sounds like the retarded will believe it, it’s BS.

Come on along,

GoingThere on October 12, 2006 at 11:33 AM

I have read the study, and 92% of those households asked to present a death certificate were able to do so. As I said i my post on teh subject a littel while ago:

Using the research of the John Hopkins study, the Iraqi Ministry of Health should be able to therefore produce roughly 602,568 total death certificates (654,965 x 92%), and 561,200 (610,000 x 92%) of these death certificates should by attributed to violent deaths, if they do in fact collect such information nationally.

I’ve emailed Omar at Iraq the Model to see if he can get this data (# of death certificates issued) from the Ministry of Health. If he can, and they’ve issued significantly less death certificates, then this study is heavily flawed.

Bob Owens on October 12, 2006 at 11:37 AM

Slublog got it right: no random

Ropera on October 12, 2006 at 11:39 AM

“prestigious” institution…
1 archaic : of, relating to, or marked by illusion, conjuring, or trickery

A new book just published might help us understand the data better: Statistics by Agenda, Dhimmi Mathematics Made Easy Allah Loves You Press 2006

GoingThere on October 12, 2006 at 11:40 AM

I’ve emailed Omar at Iraq the Model to see if he can get this data (# of death certificates issued) from the Ministry of Health. If he can, and they’ve issued significantly less death certificates, then this study is heavily flawed.

Bob Owens on October 12, 2006 at 11:37 AM

Or the Ministry of Health is heavily flawed. I believe that about 83% of HH were asked for a death c, and 92% were able to comply, hence the 75%. That’s from memory, could be wrong.

honora on October 12, 2006 at 11:41 AM

Nonetheless (no disrespect intended) I would need more evidence that the method yielded bad data, being somewhat familiar with the reputation of both institutions.

For some reason, Lancet has not released an appendix (or I have not seen it) with the survey instrument as part of it. Knowing which questions were asked would help determine part of that, I think.

Personally, I think the sample was flawed, and that’s the reason for the poor results.

Slublog on October 12, 2006 at 11:41 AM

It seems the motive of the Lancet study was to come up with a ridiculous number (652,000), have all the network news outlets parade the number as fact right before an election by giving it enormous press coverage, and then when serious researchers (including those in Iraq) debunk the number, it gets no press coverage at all.

Now everybody who gets their news from watching just thirty minutes in the evening believes that 652,000 people have died in the war. This Lancet study was pure propaganda.

And further, where are the bodies?

Also, where was the press coverage of all these people dying? You would think we would have heard about it by now, considering how liberals in the MSM relish reporting when American soldiers and civilians die, so they can turn public sentiment against the war. Remember how we heard about every single “innocent civilian” dying in Lebanon at the hands of the Israelis?

januarius on October 12, 2006 at 11:45 AM

If it sounds to good to be true it usually isn’t. A South Park extension to that axiom, if the retarded will believe it, it is total BS.

I cannot think of anyone that would want this to be actually true (except Iran), but I can think of a whole bunch of people that would want it to be thought of as true.

GoingThere on October 12, 2006 at 11:48 AM

Nonetheless (no disrespect intended) I would need more evidence that the method yielded bad data, being somewhat familiar with the reputation of both institutions.

For some reason, Lancet has not released an appendix (or I have not seen it) with the survey instrument as part of it. Knowing which questions were asked would help determine part of that, I think.

Personally, I think the sample was flawed, and that’s the reason for the poor results.

Slublog on October 12, 2006 at 11:41 AM

I guess the definition of household could be an issue, particularly in a wartorn country with a cultural tendency towards extended family units. My husband subscribes to a service (Muse I think it’s called, it’s a JH service) that downloads medical journal articles in a more complet format than is available to the public. I’ll check this evening and see if there’s any more info available.)

Given the range of responses–someone is either dead or they’re not–the sample size relative to universe size can be pretty modest so you’re right, the questions and the sampling techniques need to be looked at. My understanding of the cluster technique does not suggest it cannot be a good predictive tool, but I am not a statistician.

honora on October 12, 2006 at 12:07 PM

Allah, why is the number “so much higher” than IBC has it? What is your reasoning? It seems to me that you are equivocating to find a reasonable number, but the number of deaths is only a small part of the issue.

Look, if the IBC reports 20K deaths and there are 20,010, that’s an error in counting or reporting. But their methodology is still not flawed. But if the Lancet reports 20,010 it’s still a flawed study that is arbitrary and capricious and leads to spurious conclusions.

What is in fact worse is the lack of context given to these numbers. The Lancet reports have consistently pointed the finger a America as part of their report (deaths due to violence up after America comes on the scene). No mention of non-uniformed combatants counting as casualties, foreign-born combatants counting as casualties, displaced people being counted as casualties “because the house is empty”.

At least the IBC uses demonstrable data. The Lancet report of 2004 took a “random” sample from demonstrably violent, urban areas and extrapolated. That would be valid only if the whole country was #1 urban, and #2 densely populated.

The truth is that using the IBC methodology, if they find 20,000 more bodies, the count goes up 20,000. If the Lancet finds 20,000 more bodies, the casualty rate would be reported in the millions.

DDG on October 12, 2006 at 12:30 PM

How can six hundred thousand people just, disappear? Isn’t the about the same number killed by Hussein over a twenty year rule? If this study is true, where are all those people?

Savage on October 12, 2006 at 12:30 PM

Haven’t read all the comments here, but I think we hashed it all out in the first thread, and I just wanted to add one thing… Aside from the fact that we would know if there were hundreds of thousands dead in this country swarming with media, and some mentioned specifically al-Jazeera, etc. But also, if it was that kind of situation, would we have the boom in reenlistment if the situation were anywhere close to what this “study” suggests?

RightWinged on October 12, 2006 at 12:35 PM

Caveat: I have not read the study but from what I have read, the methodoly involved asking HH for deaths in their HH and getting death certificates for about 75% of reported deaths.

Which would tell us that 3 in 4 deaths were reported and death certificates were issued. But according to the Lancet numbers vs. the Iraqi governmemnt numbers, if we’re to believe Lancet’s then 11 in 12 deaths went unreported and were not documented. It doesn’t add up.

Pablo on October 12, 2006 at 12:37 PM

Which would tell us that 3 in 4 deaths were reported and death certificates were issued. But according to the Lancet numbers vs. the Iraqi governmemnt numbers, if we’re to believe Lancet’s then 11 in 12 deaths went unreported and were not documented. It doesn’t add up.

Pablo on October 12, 2006 at 12:37 PM

Well I am willing to be convinced the Lancet made a huge booboo; I am equally willing to believe that the Iraqi “government” is disfunctional and therefore look with some suspicion on their numbers. (Look at it this way: who is more motivated to lie?)

I am torn on this–on one level it’s hard to believe the Lancet and JH would do shoddy work; on the other it’s hard to believe this huge number of people are dead. We’ll see.

honora on October 12, 2006 at 12:44 PM

on one level it’s hard to believe the Lancet and JH would do shoddy work…

Just look at the last version of this report they churned out. This one is actually a methodological improvement from the last.

Pablo on October 12, 2006 at 12:56 PM

Just look at the last version of this report they churned out. This one is actually a methodological improvement from the last.

Pablo on October 12, 2006 at 12:56 PM

Do you expertise in this area? I ask because your statement 3 posts above makes no sense.

The correct conclusion would be: if Lancet is wrong: the sample size or reporting technique is corrupt OR if Lancet is right: the Ministry doesn’t keep count of the death certificates they issue. (Remmber, Lancet says 75% of the people they reported dead have associated certificates. So according to them, 75% of deaths were reported And based on that and the sample size, they make a projection. The point being, they are not basing their claim on the notion that 11 in 12 deaths go unreported.)

As for the government numbers? Well are these the same guys that can’t get the electricity back on?

We’ll see.

honora on October 12, 2006 at 1:33 PM

It seems the motive of the Lancet study was to come up with a ridiculous number (652,000), have all the network news outlets parade the number as fact right before an election by giving it enormous press coverage

I think every logical person here realizes this is the case. Just look at the dozens upon dozens of “scandals” the Dems/media have created over the past few years, slowly chipping away at GOP, War on Terror, Iraq, support. It’s all about getting Dems back in power and nothing more. Each time one of these “scandals” comes out, the first reports are all that matters. In the end there is almost never any wrong doing, but the seed has been planted and even on the rare occassion when the public gets the message, that there was no scandal, the mystery surrounding it makes people uneasy anyway. The Dems understand this all to well. I heard it put once – “If you go to the doctor and he says you have a tumor, slices you open to get it out and finds out he made a mistake and there was no tumor, you’re still stuck with that scar. Sure you’re glad you don’t have a tumor, but you will never get rid of that scar”.

RightWinged on October 12, 2006 at 2:16 PM

I would agree that the numbers seem outrageously high – probably too high to be accurate. I’d be surprised if the number were anywhere north of 250,000.

But is 250,000 even an acceptable number? That’s a LOT of people. Surely not all of them “had it coming to them”. What percentage of them are civilians?

What is an acceptable number of dead? Where is the point that it is no longer a “just” war, and becomes a slaughterhouse?

I’m sure that most of you don’t think we have reached the slaughterhouse point yet, but is there a point when enough is enough? And if so, where is that number?

Surely not everybody in here echo’s Gregor’s sentiment of

the higher the number, the better.

GregH on October 12, 2006 at 2:17 PM

I am in a state of confusion. how do we not know how many people are dying?

Hey man I just put 30 bodies in the ground today that makes uh 400 this month crap that is alot!

The MSM itself is orgasming over every body that isn’t breathing read the friggin’ headlines over the last year and add the nuber from the friggin’ head line for crissakes’. The wheels are’t coming off they have friggin’ forgot what wheels are!

This pisses me off so much I might count the headlines. Isn’t there a website that is counting the number of people killed by radical islam around the world since 9/11. My memory sucks but I don’t think their total is that high!

Priest on October 12, 2006 at 2:27 PM

GregH, anything above zero

Entelechy on October 12, 2006 at 2:47 PM

Correction:

GregH, anyting, below zero

Entelechy on October 12, 2006 at 2:48 PM

I would agree that the numbers seem outrageously high – probably too high to be accurate. I’d be surprised if the number were anywhere north of 250,000.

But is 250,000 even an acceptable number? That’s a LOT of people. Surely not all of them “had it coming to them”. What percentage of them are civilians?

What is an acceptable number of dead? Where is the point that it is no longer a “just” war, and becomes a slaughterhouse?

I’m sure that most of you don’t think we have reached the slaughterhouse point yet, but is there a point when enough is enough? And if so, where is that number?

Surely not everybody in here echo’s Gregor’s sentiment of

the higher the number, the better.
GregH on October 12, 2006 at 2:17 PM

I think it was Stalin who said “10 dead men are a tragedy, 10,000 dead men are a statistic”. Or something like that. We lack the capacity to grasp what this really means in terms of flesh and blood, somebody’s Dad, somebody’s son.

honora on October 12, 2006 at 2:50 PM

From the HA thread on Bozell/Turner:

This idea of an antiseptic war is the pipedream of desk jockeys and politicians. Until we get the stomach of our forefathers for the violence germane to war and take this fight outside the politically correct box it has been put in by the international community we will continue to be our own worst enemy.

Alden Pyle on October 12, 2006 at 9:55 AM

Entelechy on October 12, 2006 at 2:59 PM

Lancet-published studies actually damaged public health.

The now throughly discredited study they ran linking MMR shots to autism, created a pervasive and on-going myth. MMR shots have now fallen to 40%

They’ve created at least one public health threat.

I trust them about as much as I trust tobacco company doctors.

Bob Owens on October 12, 2006 at 3:02 PM

That should have read “fallen to 60% in parts of Britain.”

oops.

Bob Owens on October 12, 2006 at 3:03 PM

For some reason, Lancet has not released an appendix (or I have not seen it) with the survey instrument as part of it. Knowing which questions were asked would help determine part of that, I think.

Personally, I think the sample was flawed, and that’s the reason for the poor results.

Slublog on October 12, 2006 at 11:41 AM

You can find the appendices on the CNN website, in the middle of the story there’s a link, couldn’t copy it for some reason.

http://www.cnn.com/2006/WORLD/meast/10/11/iraq.deaths/index.html

honora on October 12, 2006 at 3:18 PM

I’m sure that most of you don’t think we have reached the slaughterhouse point yet, but is there a point when enough is enough? And if so, where is that number?

GregH, what’s your solution? We aren’t killing most of those people. Do you think if we were to say “we’ve reached the slaughterhouse point” and leave things wouldn’t get worse? Smarten up. By the way, I don’t accept your premise in the first place because I believe the number is much closer to 50k.

RightWinged on October 12, 2006 at 3:24 PM

Surely not all of them “had it coming to them”.

How do you know that to be true? Are you just guessing? Assuming? What are you basing that on? Have you spent time in Iraq during the war? Do you personally know Iraqi dead who you know were sweet innocent victims?

Whatever the number of dead in Iraq might be, it is my “opinion” that the overwhelming majority were PROBABLY the enemy, which means they “had it coming.”

Gregor on October 12, 2006 at 4:31 PM

Greg H –

Which is preferable?

600,000 dead sadistic bloodthirsty jihadists?

or

20,000 dead sadistic bloodthirsty jihadists?

And for the purpose of this question you’re not allowed to answer with some French theory that it would be preferable to offer more dialogue and negotiation in place of any dead.

Just answer the question with the choices given.

Gregor on October 12, 2006 at 5:45 PM

Umm, let me see here … 655,000 dead. The invasion started roughly 3 and half years ago, plus almost one month, so …

365 (days in a year) times 3.5 years = 1277.5, plus … ah heck, lets just say 1300 days. Now …

655000/1300 = 503.846153846153846…

500 deaths per day. Every day. For three and a half years. Plus almost a month.

And the MSM, which wets itself when a bomb kills 20+ in one swoop, didn’t say anything about this???

I call bullshit. Long, loud and clear.

LissaKay on October 12, 2006 at 10:52 PM

test

Ropera on October 13, 2006 at 12:58 AM

Do you expertise in this area? I ask because your statement 3 posts above makes no sense.

The correct conclusion would be: if Lancet is wrong: the sample size or reporting technique is corrupt OR if Lancet is right: the Ministry doesn’t keep count of the death certificates they issue. (Remmber, Lancet says 75% of the people they reported dead have associated certificates. So according to them, 75% of deaths were reported And based on that and the sample size, they make a projection. The point being, they are not basing their claim on the notion that 11 in 12 deaths go unreported.)

honora, is there any evidence that the Ministry can’t/won’t count, or is this entirely conjecture on your part? That seems to be the point of your post. Otherwise, the extrapolation they should have made regarding DCs is that there about 400,000 of them missing.

As for the government numbers? Well are these the same guys that can’t get the electricity back on?

You realize that that’s a silly argument, don’t you?

Pablo on October 13, 2006 at 6:46 AM

Pablo,

Not only that, but during the time this “study” was compiled, Iraqi morgues were free to give their mortality figures directly to the press… and they did.

Funny how those hundreds of extra bodies a day just weren’t noticed…

Bob Owens on October 13, 2006 at 9:08 AM

Bob,

Funny how those hundreds of extra bodies a day just weren’t noticed…

Right, because door to door interviews with Iraqis are the path to truthiness, whereas the Iraqis who actually count the dead bodies and record the deaths must somehow be lying, as they want to keep the true picture from the rest of the world. Or, they’re just too incompetent to add the figures.

It makes no sense.

Pablo on October 13, 2006 at 9:41 AM

honora, is there any evidence that the Ministry can’t/won’t count, or is this entirely conjecture on your part? That seems to be the point of your post. Otherwise, the extrapolation they should have made regarding DCs is that there about 400,000 of them missing.

As for the government numbers? Well are these the same guys that can’t get the electricity back on?
You realize that that’s a silly argument, don’t you?

Pablo

Re-read your message.

Which would tell us that 3 in 4 deaths were reported and death certificates were issued. But according to the Lancet numbers vs. the Iraqi governmemnt numbers, if we’re to believe Lancet’s then 11 in 12 deaths went unreported and were not documented. It doesn’t add up.

The whole Lancet study is underpinned by the claim that 75% of the people they claim are dead are associated with a death certificate. So 3 of 4 within the sample population have a death certificate–which contradicts your statement that 11 in 12 deaths went unreported

per Lancet.

If their projections are right, then there is shortage of death certificates versus the number of dead.

You’re taken the sample info and gov’t info and assigned each to the wrong source.

And my argument re the electricity? It’s not only not silly, it cuts right to the heart of the matter. Don’t get me wrong, I suspect the Lancet numbers are wrong (or maybe I want them to be wrong). But the notion that the Iraqi “government” completely screws up something like this, no problem at all believing that.

honora on October 13, 2006 at 10:41 AM

The whole Lancet study is underpinned by the claim that 75% of the people they claim are dead are associated with a death certificate. So 3 of 4 within the sample population have a death certificate–which contradicts your statement that 11 in 12 deaths went unreported

No, 3 in 4 of their sample had a DC. NOT the people they claim died.

From looking at the reported numbers and looking at the difference between those and Lancet’s numbers, if we assume that Lancet’s numbers are correct, then we must also assume that only 1 death in 11 was recorded because they only have records for 1/12 of the deaths Lancet projects. Or, that the Iraiqs can’t/won’t count and report them, which you seem to be leaning toward.

And yet, they tell uf that a representative sample can produce a DC in 75% of cases. So, we can’t find the bodies, we can’t find the records, and we can’t find a trend that says the vast majority of reaths aren’t recorded. This is not complicated.

But the notion that the Iraqi “government” completely screws up something like this, no problem at all believing that.

You think they’re incapable of basic record keeping? Why?

Pablo on October 13, 2006 at 12:34 PM

I’ve read the study.

The key to understanding this study comes in this sentence:

Sampling followed the same approach used in
2004,8

Their methodology (picking families via geography) allows them to CHERRY PICK their subject families and then claim the results are representative of the entire nation when extrapolated. That was the fault of the 2004 study and is the problem with it now.

I can not prove with certainty that the selection of families (and the neighborhoods they live in) was deliberately designed to result in a pre-selected outcome. I suspect that the neighborhoods were selected to obtain the highest estimates. I suspect this the case given who the authors and the Lancet are and the Lancet’s political point of view.

In addition, unlike other studies in the Lancet, this study does not appear to be refereed. In fact, one of the authors, Les Roberts — who is credited as being the instigator of it — is openly outspoken against the war and is openly left wing. His interview in the Socialst Worker alone ought to disqualify this study as having any validity, especially as the study itself piously and self-righteously claims it has no “conflictis of interest” — I the Lancet is expecting the public to ignore the POLITICAL PREJUDICES of the author and pretend that this isn’t a politically inspired document.

Therefore, the conclusions of this study are wrong as it is an example of circular reasoning bolstered by cherry picked data.

Or as Mark Twain was reputed to have said: There are lies, damned lies, and statistics.

This is not the first time that statistical studies have been used to achieve a political end. Almost all gun control studies come with built in biases designed to deliver a desired result, for example. The infamous “43 to 1” study by Kellerman, et al (1993) study in the New England Journal of Medicine being a specific example.

And as about 70% of “studies” reported in the press concerning actual medical topics are contradictory and fall into the realm of “junk science,” this study deserves that appellation as well.

In otherwords, this study is BULLSHIT!

This study is a POLITICAL STATEMENT not a scientific one.

As I noted in a previous post in another thread, the anti-war movement can only succeed if they strike from a “high moral” ground, the highest being human lives lost.

In spite of media hysteria and inflamed rhetoric, American KIA and WIA loses in Iraq and Afghanistan are lower that at any other time in our history, proportionate to the personnel deployed. This result is not an accident given that the Military has labored to increase its efficiency since the All-Volunteer Military program began in the 1970’s. Most Americans understand this, which is why they have achieved almost no traction with the argument.

Therefore, in their attempt to “shame us” into cutting and running in Iraq, the leftists are now reduced yapping about lost IRAQI lives.

Bluntly, I personally do not give a rat’s ass about the exact number of Iraq casualties, as even the study noted that most of them are RED ON RED (caused by other Iraqis).

So, I’ll be a little more blunt and simply say: It’s too damn bad that Iraqis are killing other Iraqis in a RELIGIOUS FEUD that’s about 1400 years old that is being waged by people with a 7th mindset. BUT IT IS NOT AMERICA’S PROBLEM!

This fued wasn’t caused by George W. Bush. It wasn’t caused by the American invasion. It wasn’t caused by the presence of American troops in Iraq, it is caused by SHIITES AND SUNNIS HATING EACH OTHER, something has existed since one bunch of religious wack-jobs murdered the leader of another bunch of wack-jobs back in the 7th Century.

That fact that America is TRYING TO QUELL this violence speaks volumes of OUR MORAL COMPASS AS A NATION and gives us the high moral ground here, not the anti-war left.

The fact that the flames of religious hatred are being fanned by Iran and that Syria’s borders are porous — both of whom are trying to defeat America in Iraq — indicates that America needs to deal with two more terrorist supporting states before this is all over.

In fact, it is inevitable that America confront and destroy the Islamofascists running Iran in order to defeat Islamofascism.

Honora, I’ll save you the trouble of asking. I have a degree in mathematics, and I know what I’m talking about.

georgej on October 15, 2006 at 1:09 PM